During their lifetime, one in every eight U.S. women will be diagnosed with breast cancer, with 182,500 newly cases in 2008. Over 57,000 breast reconstruction operations are performed in the U.S. annually. Breast reconstruction can be performed immediately after mastectomy or delayed months to years later. 60% of the reconstructions in the U.S. involve tissue expanders followed by exchange for a permanent implant. Autologous flap breast reconstruction is much less common . There are no randomized trials to determine the comparative effectivess of immediate vs. delayed breast reconstruction. Decisions about the type and timing of reconstruction are based on individual surgeon and patient preference. Breast reconstruction complication rates and outcomes have been reported primarily from small studies, and almost exclusively from single institutions. Comprehensive multicenter studies to determine rates of surgical site infection and other wound complications after immediate vs. delayed reconstruction have not been performed. In single-center studies, surgical site infection rates are signficantly higher after immediate reconstruction compared to mastectomy only. In cases with an implant, 2/3 of infected implants are lost. We propose to use geographically diverse longitudinal claims data from the largest commercially insured population in the U.S. to determine the incidence of surgical site infection and noninfectious wound complications after mastectomy and immediate vs. delayed reconstruction. We will determine factors associated with risk of infectious and noninfectious wound complications in women who had mastectomy with or without immediate reconstruction from 2004-2007. We will use these data to develop risk prediction models for wound complications, taking into account the specific type of breast reconstructive surgery. We will then validate the risk prediction models with claims data from later years. These results will help determine if all of a subset of women could potentially benefit from delayed reconstruction, and if there are specific processes of care that protect against wound complications. This study will provide significant new information which can be used to improve outcomes for women with breast cancer.